Some ideas … Have you checked that the feasible region is not empty (e.g. ramp rate setting Pmax less than Pmin)? Does it solve if you set RATE_A = 0 for all branches to remove line constraints? Can you try a different solver? Do you get a NaN issue with other solvers as well? If necessary, maybe you can use the Matlab debugger to track down where the NaN is occurring.
Ray > On Nov 23, 2015, at 1:46 AM, Kalin Lee <[email protected]> wrote: > > I am implementing must-run variable generation and ramp rate limits, but with > essentially limitless (more expensive) unconstrained generation. But during > some time steps I get non-convergence (Error Below). Does anyone have any > insight into this issue? > > For must-run I set the generator Pmax and Pmin to the generation output. > For the ramp (up) rate limit I set the generator Pmax to the previous output > + X MW. > > > Error Below: > > The interior-point algorithm uses a built-in starting point; ignoring > user-supplied X0. > Exiting: cannot converge because the primal residual, dual residual, or > upper-bound feasibility is NaN. > Index exceeds matrix dimensions. > > Error in qps_ot (line 272) > mu_u(ilt) = lam.ineqlin(1:nlt); > > Error in qps_matpower (line 259) > [x, f, eflag, output, lambda] = ... > > Error in dcopf_solver (line 227) > [x, f, info, output, lambda] = qps_matpower(HH, CC, A, l, u, xmin, xmax, x0, > opt); > > Error in opf_execute (line 74) > [results, success, raw] = dcopf_solver(om, mpopt); > > Error in opf (line 226) > [results, success, raw] = opf_execute(om, mpopt); > > Error in dcopf (line 8) > [varargout{1:nargout}] = opf(mpc, mpopt); > Error in ProjectFinal (line 135) > results(i,1) = dcopf(pscase); >
